UPS impairment leads or contributes subsequently to cardiac hypertrophy

busy in these old mice. Interestingly, in a polyglutamine disease model crossed with Ub^G76V-GFP mice, a significant increase in reporter protein was also observed late in disease (Bowman et al., 2005). But the basis for this increase was explained by a corresponding increase in Ub^G76V-GFP mRNA. An in vitro assay showed normal (slightly increased) 20S activity, which confirmed that the increase in reporter protein did not result from a reduction of proteasome activity. Thus, the expression of Ub^G76V-GFP at transcriptional level should be determined in the old KI and KO mice crossed with Ub^G76V-GFP mice.

In conclusion, the increases in 20S proteasome activities and in the steady-state level of ubiquitinated proteins over the entire life span of both KI and KO mice provided clear evidence that the UPS is altered in the hearts of both mouse lines. Most of these alterations appear to be part of the general pathology related to the massive cardiac

comparison, despite normal levels of endogenous cMyBP-C, significant changes in the structure and ultrastructure of the heart were evident in the MyBP-C.mut2 mice lacking the myosin binding site (Yang et al., 1999). These changes were likely caused by the presence of the truncated protein, although it was expressed at a low level.

The M7t-TG mice also did not exhibit significant cardiac abnormalities when analyzed by transthoracic echocardiography. Neither wall-thickening nor dilatation of the left ventricle nor impaired cardiac function (determined by the FS) was found up to the age of two years. A really striking observation was the decreased VW/BW (-14%) in the M7t-TG mice at the age of 9 mo. The significance was caused by the male mice, which exhibited a highly significant decrease of the VW/BW (-18%), whereas the female mice just showed a tendency towards a decrease (-9%). The lower VW/BW did not result from a higher BW in the M7t-TG mice, but indeed from a lower VW. Interestingly, transgenic mice expressing a truncated cardiac troponin T (cTnT) also exhibited a decreased HW/BW (-23%) vs. WT mice (Tardiff et al., 1998).

Full dissection of the heart and determination of the individual chamber weight-to-BW-ratios in cTnT mice showed that the decrease in heart mass was restricted to the left ventricle. Surprisingly, a small, but significant increase in atrial size was observed.

The decrease in cardiac mass was due to a primary loss of cardiomyocytes and a decrease in cardiomyocyte cell size. These examinations were not performed in the M7t-TG mice, but should be done in the future to identify the reason for the decreased cardiac mass. Actually, nine independent founders were generated with the cTnT-truncation construct, but only three of them expressed the transgene at the protein level, and even then at low levels (<5%). Northern blot analysis showed that the transgene was expressed at high mRNA levels. These findings are suggestive of UPS involvement.

Immunofluorescence analysis of AMVM indicated a normal sarcomeric organization in the KI mice. However, the KI mice exhibited myocyte and eccentric LV

1999). At 2-3 mo of age histological abnormalities including myocyte hypertrophy, myofibrillar disarray, interstitial fibrosis and dystrophic calcification were observed in the cMyBP-C^t/t mice. In contrast, the Het mice did not develop LV hypertrophy up to the age of 18 mo, but exhibited slight myocyte hypertrophy and a higher, not significant level of β-MHC mRNA at the age of 3 mo. In comparison, at the age of 12 mo only 32% of the cMyBP-C^t/+ mice had developed a cardiac hypertrophy (McConnell et al., 2001). In >30-mo-old cMyBP-C^t/+ mice, LVAW and LVPW were significantly greater than in WT mice, and significant increases in hypertrophic markers (ANP and BNP) were found. Histological sections derived from >30-mo-old cMyBP-C^t/+ mice demonstrated myocyte hypertrophy, interstitial fibrosis and myofibrillar disarray in 50% of mutant animals. However, cardiac function in >30-mo-old cMyBP-C^t/+ mice was indistinguishable from that of respective WT mice.

Taken together, the hypothesis that alterations in the UPS contribute to the development of hypertrophy in cMyBP-C transgenic mice could not be substantiated.

In case of KI and KO mice, this is likely due to the massive cardiac phenotype due to the complete absence (KO) or large reduction (KI) in cMyBP-C protein. The absence/reduction in cMyBP-C protein by itself is likely a strong trigger for hypertrophy, overriding any further contribution by other factors. It should be mentioned in this respect that it is quite surprising that the KO mice survive the absence of one of the major sarcormeric proteins at all. Moreover, one may argue that the similarly severe phenotype in KI mice, expressing at least 10% of missense cMyBP-C could be related to the affection of the UPS. But this idea needs further studies. On the other hand, the atrophy in M7t-TG mice could argue for just the opposite, namely that the stimulation of the proteasomal degradation machinery in 9-mo-old animals was causally related to atrophy. But this remains speculative at this point.

NMD is involved in the regulation of cMyBP-C mutant levels in KI and Het mice.

Characterization of the cMyBP-C knock-in mouse model revealed that the level of total cMyBP-C mature-mRNA, but not of the pre-mRNA, was markedly reduced in

level of processing pre-mRNAs to mature-mRNAs. A known pathway that specifically eliminates nonsense mRNAs is the NMD. NMD is activated in mRNAs that contain a PTC located >50-55 nt upstream of the last exon-exon junction (Nagy and Maquat, 1998). In the KI and Het mice, the PTC is located in exon 9 in the nonsense mRNA and therefore far upstream of the last junction between exons 34-35 suggesting an involvement of the NMD. Indeed, two differently acting NMD inhibitors, CHX and emetine, markedly increased the level of the nonsense mRNA, but not of the missense mRNA in KI NMCM. This specificity for the nonsense mRNA fits with the rule that NMD only targets aberrant mRNAs containing a PTC.

More importantly, CHX treatment in vivo resulted as well in increased levels of nonsense mRNA in the KI and Het mice indicating that NMD is operational in the intact whole animal heart. The reduction in missense mRNA points to an additional mechanism, which is not yet known. As already mentioned above, similar expression levels of mutant cMyBP-C transcripts as in the KI and Het mice were obtained in the cMyBP-C^t/t and cMyBP-C^t/+ mice. The presence of a PTC together with the markedly reduced levels of total cMyBP-C mRNA suggest a NMD involvement in these mice like in the KI and Het mice.

Taken together, an important new result from these studies is that NMD participates in the regulation of pathological cMyBP-C transcripts in the whole animal. On the basis of the results obtained previously (Sarikas et al., 2005) a model can be proposed that involves a two-step quality control. Transcripts are degraded by the NMD if they comply with the NMD rules (PTC >50-55 nt upstream of the last exon-exon junction) and the UPS comes into play if truncated proteins are being produced. The latter is the main degradation system in cases where a cDNA is overexpressed that is devoid of an exon-intron structure.

The lack of cMyBP-C is probably the main cause of hypertrophy, which in turn

WT mice and probably therefore these mice did not develop cardiac hypertrophy up to the age of two years. The Het mice expressed 79% full-length cMyBP-C (mutant and normal) compared to WT mice and did not develop cardiac hypertrophy up to 18 mo of age. But the reduced level of cMyBP-C protein together with myocyte hypertrophy, the increased (not significant) level of β-MHC mRNA at 3 mo of age and the published data in the >30-mo-old cMyBP-C^t/+ mice suggest that myocyte hypertrophy precedes LV hypertrophy and that the Het mice will develop cardiac hypertrophy later in life. Therefore, examination of >18 mo-old Het mice is absolutely necessary.

In contrast to the M7t-TG and Het mice, KI and KO mice did not express any normal cMyBP-C and exhibited an increased HW/BW already at birth suggesting lack of cMyBP-C as the disease mechanism leading to hypertrophy. The KI mice expressed of course 10% full-length E256K protein, but whether the E256K mutant has a function at all remains at the moment elusive.

If the sarcomere function would be altered or impaired due the insufficient amount of cMyBP-C, the heart may try to compensate the functional deficits by increased myocyte growth resulting in cardiac hypertrophy. When cardiac function remains inadequate despite growth, other pathways are likely activated, as evidenced by an altered gene expression (ANP, BNP, ß-MHC), myocyte death (apoptosis) and fibrosis.

In general, the UPS is involved in a broad range of cellular pathways, such as apoptosis (Breitschopf et al., 2000; Li and Dou, 2000), cell cycle (King et al., 1996), cell differentiation (Helin, 1998) and immune and stress response (Palombella et al., 1994; Rock et al., 1994). In addition, a number of key regulatory pathways that promote cardiac hypertrophy are either targets or components of the UPS. For instance, it has been reported that several signaling proteins such as β-catenin and calcineurin, which mediate cardiac growth (including pathological hypertrophy), are degraded by the UPS (Glickman and Ciechanover, 2002; Li et al., 2004). Thus, the UPS was likely activated in response to hypertrophy in KI and KO mice and the increase in proteolytic activities was likely a compensatory mechanism against further development of hypertrophy: the 20S proteasome worked more to get rid of the hypertrophic, hypertrophy-promoting and proapoptotic proteins. This was also indicated by the positive correlation between chymotryspin-like activity and

Thus, these data as well as several recent other studies (Depre et al., 2006; Razeghi et al., 2006) are compatible with the idea that cardiac hypertrophy is generally accompanied by alterations in the UPS, mainly in the sense of increased turnover rate.

Our observation that the steady-state level of ubiquitinated proteins did not correlate with the degree of cardiac hypertrophy (in both KI and KO mice) argues for the notion that the system normally reacts to increased protein load by increased activity to keep the steady-state level constant. Accordingly, a negative correlation existed between chymotrypsin activity and the level of ubiquitinated proteins in the KO mice.

In other words, the better the proteasome works, the lower the steady-state level of ubiquitinated proteins. Interestingly, this correlation was absent in KI mice, suggesting a specific defect as discussed above. When interpreting these data, it is important to keep in mind that the tests for proteasome activity are performed in vitro, i.e. in the absence of any potentially inhibiting truncated protein, whereas the steady-state level of ubiquitinated proteins is determined in the in vivo situation.

Outlook

For the future, some investigations and treatments are yet outstanding for the M7t-TG, KI and Het mice. First, further experiments should be designed to stabilize and detect the truncated protein. In the M7t-TG mice, the UPS is likely involved in the rapid degradation of the truncated protein, although it could not be substantiated under the used conditions. The i.v.-injection of MG132 or epoxomicin in adult (older than 2 mo) M7t-TG mice followed by Western blot analysis of myofilament protein preparations may result in the detection of the truncated protein. In contrast, NMD inhibition is likely useless in the M7t-TG mice: NMD is likely not responsible for the low levels of transgenic mRNA, because it requires at least 1 intron (Zhang et al., 1998) and is therefore inactive towards transgenically expressed cDNAs. A cross between the M7t-TG and KO mice would result in mice, which express only one functional cMyBP-C

the heterozygous state for cMyBP-C). To detect the truncated protein in the KI and Het mice, the best approach may be to inhibit NMD, e.g. by siRNA-mediated knockdown of SMG-1 or Upf1 (NMD key proteins), followed by blockage of the proteasome by i.v.-injection of MG132 and final analysis of myofilament protein preparations by Western blot. This strategy is only useful to detect the truncated protein, but it will not rescue the phenotype in the KI mice. As already mentioned, the truncated protein is not only lacking the myosin and titin binding site but also the MyBP-C motif with the 3 phosphorylation sites, and is therefore probably not able to play a beneficial structural or functional role. In contrast to NMD inhibition, the treatment with gentamicin, an aminoglycoside, or PTC124 could be useful to rescue the mutant phenotype in the KI mice. Suppression of the PTC in the nonsense mRNA would lead to a mutant almost full-length cMyBP-C protein. Whether this protein would have structural and functional properties at all is difficult to say, because it is at least missing the exon 6, which leads to a frameshift.

A second point, which should be investigated in the future, is the UPS function in the Het mice. The Het mice did not develop cardiac hypertrophy up to 18 mo of age and therefore no altered UPS function is expected. But maybe there is already a tendency towards an increase in both ubiquitination and degradation at this age, which proceeds with age.

Finally, it would be interesting to investigate the stress response in the M7t-TG and Het mice. A striking characteristic of FHC in humans is its preference to cause sudden death, oftentimes during significant cardiac stress brought on by exercise (Maron et al., 1980; Maron et al., 1996). But in general, the laboratory mouse is a sedentary animal in a minimally stressful environment. Controlled regimens have been therefore established to stress mice physically or pharmacologically: mice can be subject to swimming (Kaplan et al., 1994; James et al., 1998) or treadmill exercise (Fewell et al., 1997; James et al., 1998) or to chronic isoproterenol infusion (Kudej et al., 1997). It is expected that phenotypes that may not be present at rest become apparent under these controlled stress regimens. For instance, although the transgenic MyBP-C.mut2 mice appeared overtly healthy, they were significantly compromised in their exercise capacity and exhibited decreased heart rates in response to the treadmill exercise

conditions of stress overt cardiovascular abnormalities might become apparent in the M7t-TG and Het mice.

Potential therapeutic approaches

When considering therapeutic interventions, it is important to keep in mind that FHC is, in the vast majority, an autosomal-dominant disease, i.e. it occurs at the heterozygous state. This is different from the mouse models for yet incompletely understood reasons.

Assuming that the (partial) absence is the main cause of hypertrophy in cMyBP-C mutations, putative therapeutic approaches should aim at increasing the amount of full-length cMyBP-C protein. An estimated 70% of the 165 known MYBPC3 mutations (Richard et al., 2006; Alcalai et al., 2007) should result in a PTC located

>50-55 nt upstream of the last exon-exon junction and therefore involve NMD. Up to date, only two papers have analyzed the consequences of frameshift MYBPC3 mutations in human myocardial tissue (Rottbauer et al., 1997; Moolman et al., 2000).

In one patient, an insertion of a single guanine resulted in a newly created splice donor site in exon 25 and a PTC in exon 26 (Moolman et al., 2000). The level of nonsense mRNA was 2-fold lower than the WT mRNA in this patient suggesting the involvement of NMD. In general, NMD is a quality control of the cell and prevents the production of nonfunctional or potentially harmful truncated proteins (Frischmeyer and Dietz, 1999; Maquat, 2004). However, some truncated proteins deleted for the C-terminal domains have been reported to support normal cell functions (Sheppard et al., 1994; Kerr et al., 2001). In the case that a mutant protein retains residual activity that can partially retain normal protein function, NMD might augment the original defects (Cali and Anderson, 1998; Ainsworth, 2005). Thus, the selective inhibition of NMD may provide a novel strategy to rescue the mutant

phenotype in fibroblasts of a patient with Ullrich disease (Usuki et al., 2004; Usuki et al., 2006). But in this context it has to be noted that it is not known what kinds of deleterious side effects may arise from a NMD inhibition in humans. For instance, otherwise innocuous, recessive PTC-related mutations carried in heterozygous condition as part of an individual’s genetic load could cause unpredictable, adverse side effects. In addition, the accumulation of natural mRNAs could cause adverse side effects.

Other studies have shown that treatment with aminoglycosides can also rescue the mutant phenotype in PTC-related diseases by allowing a PTC read-through, thus permitting the synthesis of a near-normal protein (Howard et al., 1996; Barton-Davis et al., 1999). Aminoglycosides bind to the decoding site of ribosomes and promote the incorporation of an amino acid at the PTC by a near-cognate aminoacylated transfer (t)RNA (Rospert et al., 2005). They have been already used in human patients to suppress PTCs within transcripts associated with cystic fibrosis (Kuzmiak and Maquat, 2006). Unfortunately, long-term use of these antibiotics in humans can cause nephrotoxicity and ototoxicity. The recent development of a new drug, PTC124, which promotes amino acid incorporation at PTCs by an unknown mechanism (Ainsworth, 2005) and which did not evoke kidney failure or deafness in phase I of safety trials, is a promising perspective.

Whether UPS inhibition is a useful strategy to prevent hypertrophy, is depending on the cMyBP-C mutation. If the cMyBP-C mutation results in a nonsense mRNA containing a PTC located <50-55 nt upstream of the last exon-exon junction and leads to the production of a truncated protein, which retains residual activity, UPS inhibition would result in an useful stabilization of this protein. The proteasome inhibitor bortezomib is already in clinical use for the therapy of multiple myeloma (Richardson and Anderson, 2003) and, until now, appears to have a favorable safety profile that does not include cardiac toxicity.